NVIDIA GeForce GTX 650 Ti OEM

The NVIDIA GeForce GTX 650 Ti OEM, built on the Kepler architecture with the GK106 chip, occupies a specific niche in the GeForce 600 generation. Produced on TSMC’s 28 nm process, this card integrates 2,540 million transistors across a 221 mm² die, resulting in a transistor density of 11.5 million per square millimeter. With a production status of end-of-life and a release date in early 2013, it represents a mature design that now sits at the 50th percentile among all GPUs in the database, indicating mid-pack standing in overall performance distribution.

Benchmark Performance

The GTX 650 Ti OEM delivers compute throughput that defines its position in the performance hierarchy. Its FP32 output reaches 1,425.4 GFLOPS, a figure that reflects the card’s 768 shading units operating at the given clock configuration. The texture rate of 59.39 GTexel/s, supported by 64 texture mapping units, allows for solid texture-heavy workload handling at moderate settings. Pixel throughput is rated at 14.85 GPixel/s, driven by 16 raster output units. These numbers, taken together, place the card in a class where 1080p gaming at medium details is feasible, though the data does not include specific game benchmarks or rival scores to quantify exact deltas. The percentile rank of 50 means that half of all GPUs in the database score higher and half score lower, positioning this card as an average performer in the broader landscape. Without nearest rivals listed, the comparison must rely on architectural context: the Kepler generation’s strengths in power efficiency and balanced shader utilization are evident in these raw throughput figures. The absence of a base or boost clock in the fact pack means the memory clock of 1350 MHz (5.4 Gbps effective) is the only timing reference, which aligns with the card’s mid-range positioning rather than high-end ambitions.

Memory Subsystem

The memory configuration consists of 2 GB of GDDR5 across a 128-bit bus, yielding a bandwidth of 86.40 GB/s. This combination is a critical constraint for high-resolution workloads. At 1080p, the 2 GB capacity is generally sufficient for textures and frame buffers in titles from the era, but the 128-bit bus width limits how quickly data can be fed to the GPU core. The 86.40 GB/s bandwidth is modest by modern standards, and benchmark data suggests that higher resolutions, such as 1440p or 4K, would strain this subsystem. Texture-heavy scenes with large, uncompressed data sets could cause memory pressure, leading to stuttering or reduced frame pacing. The memory clock of 1350 MHz, translating to 5.4 Gbps effective, is the sole clock figure provided, and it reflects the card’s intended balance between cost and capability. For users targeting maximum settings in newer titles, the bus width and bandwidth would become the limiting factor, even if the shading units could theoretically handle more complex geometry. In contrast, older or less demanding games that fit within the 2 GB frame buffer would see adequate performance, but the data does not support claims of headroom beyond that scope.

Ray Tracing and Feature Set

The GTX 650 Ti OEM does not include dedicated ray tracing cores or tensor cores, as these are absent from the fact pack. Its feature set relies on the Kepler architecture’s standard capabilities, which predate hardware-accelerated ray tracing. API support includes DirectX 12 (11_0), OpenGL 4.6, and Vulkan 1.2.175. The DirectX 12 (11_0) designation indicates that while the card can run DirectX 12 applications, it operates at the feature level 11_0, meaning it lacks the higher-tier features like bindless resources or conservative rasterization that newer hardware supports. This is a significant limitation for modern titles that leverage those advanced features. Vulkan 1.2.175 support provides a pathway for cross-platform rendering, but without ray tracing or tensor operations, any AI-accelerated or path-traced effects are off the table. The card’s display outputs include 2x DVI and 1x mini-HDMI 1.4a, which limits connectivity to older monitor standards; HDMI 1.4a does not support 4K at high refresh rates, capping output at 4K 30 Hz or 1080p at higher rates depending on the interface. The absence of RT and tensor cores means that any comparison to modern GPUs must focus purely on rasterization performance, where the card’s 1,425.4 GFLOPS of FP32 compute is the primary metric. The feature set is functional for its era but lacks the specialized hardware that defines contemporary graphics cards.

How It Compares

The fact pack lists no nearest rivals, which means a direct score-based comparison is not possible from the data provided. However, the percentile rank of 50 provides a baseline: this card performs at the median of all GPUs in the database, indicating that it would be outclassed by higher-percentile parts while outperforming lower-tier options. In the absence of specific rival scores, the comparison must be framed through the lens of its own specifications. The 2 GB memory and 128-bit bus are typical of entry-level to mid-range cards of its generation, but the 768 shading units give it a compute advantage over simpler designs. The 28 nm process node and 110 W TDP suggest a power-aware design, but no wattage figures for competitors are available to contrast. The suggested PSU of 300 W and single 6-pin power connector indicate that the card targets systems with modest power supplies, reinforcing its position as an upgrade path for older machines. The single-slot form factor and 145 mm length (5.7 inches) make it suitable for compact cases, a practical advantage over larger cards. Without rival data, the conclusion is that this card sits in a crowded mid-range segment, where its performance is adequate for its time but not exceptional.

Who Should Consider It

Given the performance metrics, the GTX 650 Ti OEM is suited for users who prioritize compatibility and size over raw power. The 110 W TDP and 300 W suggested PSU mean it can be installed in pre-built systems with modest power supplies, provided they have a single 6-pin connector. The 145 mm length (5.7 inches) and single-slot design allow for installation in small form factor cases where larger cards would not fit. For gaming, the data indicates that 1080p at medium settings is the realistic ceiling, with the 2 GB VRAM and 86.40 GB/s bandwidth handling older titles or less demanding esports games without issue. High-resolution gaming at 1440p or above would likely exceed the memory subsystem’s capabilities, leading to reduced performance. The lack of ray tracing and tensor cores means that any game relying on those features is not playable with hardware acceleration, and the DirectX 12 (11_0) feature level limits access to the latest rendering techniques. Users with a collection of games from the early 2010s or those who need a basic display output for productivity tasks would find this card functional. The 2x DVI and 1x mini-HDMI 1.4a outputs support dual-monitor setups, though 4K output is restricted to 30 Hz via HDMI. The 50th percentile rank confirms that while it is not a high-performance part, it is not the weakest option either. For anyone seeking modern features or high-fidelity visuals, this card falls short, but for a legacy system or a secondary machine, it remains a viable, if unremarkable, choice.